Co-funded by the European Union’s Horizon 2020 research and innovation programme under Grant Agreement no. 691797 Innovative large-scale energy storage technologies and Power-to-Gas concepts after optimization Analysis on future technology options and on techno-economic optimization Due Date 28 February 2019 (M36) Deliverable Number D7.7 WP Number WP7: Reducing Barriers Responsible Robert Tichler, EIL Author(s) Andreas Zauner, Hans Böhm, Daniel C. Rosenfeld, Robert Tichler Reviewer Steffen Schirrmeister, TKIS Status Started / Draft / Consolidated / Review / Approved / Submitted / Ac- cepted by the EC / Rework Dissemination level PU Public PP Restricted to other programme participants (including the Commission Services) RE Restricted to a group specified by the consortium (including the Commission Services) CO Confidential, only for members of the consortium (including the Commission Services)
89
Embed
Innovative large-scale energy storage technologies and ... · Innovative large-scale energy storage technologies and Power-to-Gas concepts ... energy transport via the existing gas
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Co-funded by the European Union’s
Horizon 2020 research and innovation
programme
under Grant Agreement no. 691797
Innovative large-scale energy storage
technologies and Power-to-Gas concepts
after optimization
Analysis on future technology options
and on techno-economic optimization
Due Date 28 February 2019 (M36)
Deliverable Number D7.7
WP Number WP7: Reducing Barriers
Responsible Robert Tichler, EIL
Author(s) Andreas Zauner, Hans Böhm, Daniel C. Rosenfeld, Robert Tichler
Reviewer Steffen Schirrmeister, TKIS
Status Started / Draft / Consolidated / Review / Approved / Submitted / Ac-
cepted by the EC / Rework
Dissemination level
PU Public
PP Restricted to other programme participants (including the Commission Services)
RE Restricted to a group specified by the consortium (including the Commission Services)
CO Confidential, only for members of the consortium (including the Commission Services)
D7.7 Analysis on future technology options and on techno-economic optimization Page 2 of 89
Document history
Version Date Author Description
1.0 2018-12-20 Andreas Zauner First draft
2.0 2019-04-01 Andreas Zauner Final version
3.0 2019-07-04 Andreas Zauner Implementation of comments done by var-ious reviewers
D7.7 Analysis on future technology options and on techno-economic optimization Page 3 of 89
Table of contents
Document history ............................................................................................................................ 2
Table of contents ............................................................................................................................ 3
D7.7 Analysis on future technology options and on techno-economic optimization Page 5 of 89
Executive Summary
An ecologically sustainable energy supply that is economically viable and socially acceptable is a
high priority in European policy. The European energy supply must be transformed due to energy-
related, social, economic, and environmental/climatic factors. The use of green gases on the basis
of renewable electrical energy (as hydrogen, synthetic methane, or alternative hydrocarbons from
hydrogen) has numerous advantages, which can significantly assist Europe in transitioning its en-
ergy system. These gases can also address major issues facing the development of renewable en-
ergy sources, including the long-term storage of fluctuating renewable electricity sources, alternative
energy transport via the existing gas infrastructure, the reduction of greenhouse gas emissions, the
need to find new renewable energy sources for mobility and industrial processes, and the increase
in local production and use. Sector coupling via power-to-gas (PtG) is fundamental to the transfor-
mation of the European energy system and a significant economic parameter. Further, the decar-
bonization of the European energy system must be seen as an opportunity to decisively boost Eu-
ropean leadership in innovative energy technology, energy-related transport technology and
services, and the application and implementation of mature, green gas-related technologies.
Since the market launch and development of PtG technology depend on, among other things, the
profitability (and thus mainly on the investment costs) of the plant, the potential cost reduction are
examined in this Deliverable D7.7. In addition to the key technological characteristics (e.g., state of
the art and future projects), new developments, technologies and materials, and potential future
fields of application are analyzed. Finally, the SNG production costs are calculated for different ap-
plications in order to demonstrate PtG’s potential.
Both main components of the PtG-technology, electrolyzer and methanation systems, show promis-
ing cost reduction behavior related to scaling effects and technological learning throughout the in-
vestigating period, see Figure 1-1 and Figure 1-2.
Figure 1-1: Cost development of electrolysis systems related to scaling effects and technological learning
D7.7 Analysis on future technology options and on techno-economic optimization Page 6 of 89
Figure 1-2: Cost development of methanation systems related to scaling effects and technological learning
The economic evaluation is based on the calculation of the specific production costs for SNG in
2020, 2030, and 2050 for a 100 MWel PtG plant for three different fields of application (PtG plant
powered by a photovoltaic power plant (PtG-PV); a PtG plant powered by a wind farm (PtG-Wind);
and a PtG plant powered by the public grid (PtG-Grid)). Additionally, the SNG production costs are
calculated for different PtG technologies, which are combinations of a AEC, PEMEC, and SOEC,
with a catalytic or biological methanation unit.
Figure 1-3 summarizes the results of all the calculations performed, providing a range of costs for
each scenario. The variety among the costs is due to the different technologies used for SNG pro-
duction. In early applications (i.e., 2020 and 2030), the sole use of surpluses, electricity from PV or
wind does not provide acceptable SNG production costs, due to the still relatively high investment
costs and the low achievable full load hours of the plant. In the PtG-Grid operating mode, the PtG
plant is connected to the public electricity grid and operates at times with the cheapest electricity
prices on the spot market. In early applications, PtG plants will need to run at high full-load hours
(> 4,000 h/a) to achieve low SNG production costs. In future, the lowest costs will be achieved with
a lower number of full-load hours (3,000 h/a) when the plant is operated only during periods with the
lowest electricity prices. However, several factors, such as the need to produce green gas, may
argue for higher full-load hours, albeit with higher SNG costs.
In general, there is little difference in SNG production costs according to the technology used,
whereby in future PtG plants with an alkaline electrolyzer will have slightly higher SNG production
costs than those with a PEM electrolyzer, and a PtG plant built with an SOEC and catalytic methana-
tion will tend to have slightly lower SNG production costs. Concerning the methanation technology
used there is hardly no difference in the SNG production costs. The lower SNG production costs of
the PtG plant with an SOEC and catalytic methanation unit can be attributed to higher system effi-
ciency. However, to achieve these very high efficiencies, the SOEC requires an additional waste
heat source, which is not available at every location. By contrast, it is assumed that waste heat can
be sold in the variants where an AEC or PEMEC is connected to a catalytic or biological methanation
unit. If waste heat cannot be sold, then the SNG costs would rise in these variants. Thus, the SOEC
variant would have the lowest SNG costs by far.
D7.7 Analysis on future technology options and on techno-economic optimization Page 7 of 89
Figure 1-3: Range of SNG production costs of a 100 MW plant in 2020, 2030 and 2050 for different scenarios
A PtG plant can be used in a variety of ways in the energy system. In most of the cases, the funda-
mental goal is the production of renewable gas. It may be reasonable (while taking the market situ-
ation for renewable gases into account) to not operate the plant with about 3,000 full-load hours in
or-der to achieve the lowest SNG production costs but, rather, to increase the output of the PtG plant
by increasing the full-load hours, although this would lead to higher SNG production costs. However,
as mentioned, excessively high full-load hours (> 5,000 h/a) leads to significantly higher SNG costs.
Incidentally, in a renewable energy-based energy system with a large share of fluctuating energy
sources, the PtG plant should be operated in such a way as to ensure that the power grid is not
additionally charged but is best supported. This can be done, for example, by converting the sur-
pluses from wind and PV produced in the summer into SNG and transferring them into the winter
months (i.e., long-term storage, sector coupling). Since power generation bottlenecks are likely in
the winter months (less electricity production form PV), leading to higher electricity prices, the PtG
plant should not be operated at these times. Thus, a continuous operation (full-load hours
> 6000 h/a) of the PtG plant is not desirable. The full-load hours for reasonable PtG plant operation
(gas production, SNG production costs, and grid suitability) are regarded to be in the range of 3,000
to 5,000, incurring costs of about 5.5 to 7.5 Cent/kWh in 2050.
The sensitivity analysis indicates that reducing SNG costs requires purchasing low-cost electricity,
maximizing plant efficiency, reducing investment costs, and in cases where the plant is connected
to a PV or wind park, building the PV or wind park in good locations with high full-load hours.
The deliverable shows that barriers and prejudices can be reduced to enable the successful imple-
mentation of PtG plants. However, the development of PtG technology is subject to fundamental
energy and climate policy decisions; thus, assumptions made about the future can change signifi-
cantly. This has a major impact on the future SNG production costs calculated in this report.
D7.7 Analysis on future technology options and on techno-economic optimization Page 8 of 89
1 Introduction
An ecologically sustainable energy supply that is economically viable and socially acceptable is a
high priority in European policy. The European energy supply must be transformed due to energy-
related, social, economic, and environmental/climatic factors. The use of green gases on the basis
of renewable electrical energy (as hydrogen, synthetic methane, or alternative hydrocarbons from
hydrogen) has numerous advantages, which can significantly assist Europe in transitioning its en-
ergy system. These gases can also address major issues facing the development of renewable en-
ergy sources, including the long-term storage of fluctuating renewable electricity sources, alternative
energy transport via the existing gas infrastructure, the reduction of greenhouse gas emissions, the
need to find new renewable energy sources for mobility and industrial processes, and the increase
in local production and use. Sector coupling via power-to-gas (PtG) is fundamental to the transfor-
mation of the European energy system and a significant economic parameter.
Central contributions of PtG to the energy system
Storage and transport solution: Seasonal fluctuations of renewable electricity generators can be
balanced through the injection and storage of energy carriers produced from renewable electricity
like hydrogen and/or synthetic methane produced from hydrogen into the existing natural gas
infrastructure.1 New power lines or a grid expansion can be substituted by shifting the transportation
of energy from the electric power grid to the natural gas grid. The advantage of energy transport
via the existing natural gas infrastructure is the high energy density of the natural gas grid. An
expansion of the natural gas network would lead to a much smaller topographical intervention than
an expansion of the electricity network, which would increase public acceptance and reduce real
estate costs.2
Infrastructure solution: In addition to power plants, central Europe’s gas infrastructure includes a
high-quality transmission and distribution grid as well as enormous capacities for gas storage in
caverns and porous reservoirs. Thus, the integration of renewable gases such as hydrogen or SNG
into the natural gas infrastructure would avoid the need for enormous stranded investments in the
existing energy infrastructure. Sectoral coupling of the electrical and gas grids via hydrogen
production (with optional methanation) would also allow the integration of biogas and thus an
increased greening of the gas sector. In other words, the long-term use of the existing gas
infrastructure will depend on the degree of integration of renewable gases. Climate and energy
policies can thus be advanced by the existing gas infrastructure, which can also be used to secure
the long-term use of this infrastructure.
Supply of all segments by renewable energy sources: Green hydrogen and therefrom produced
renewable hydrocarbons such as methane can be used in all energy segments (e.g., process heat,
mobility, space heating, and electrical energy), and thus foster the greening of the European energy
system. In addition to battery-based electric mobility, the use of green hydrogen or methane from
PtG plants will significantly accelerate the transition to a sustainable transport system with low or
no emissions. Hydrogen and hydrogen-based synthetic methane can be used in combustion
engines and fuel cells, and they have a strong potential to reduce primary energy input, emissions
of air pollutants (e.g., particulates and NOX), and greenhouse gas emissions. Besides its utilization
for energy production, hydrogen as a renewable resource is also important for manufacturing
1 Therefore, refer to R. Tichler, J. Lindorfer, C. Friedl, G. Reiter, H. Steinmüller (2014) FTI-Roadmap Power-to-Gas für Österreich, Energieinstitut an der JKU Linz. Herausgeber: bmvit, Schriftenreihe 50/2014. 2 Therefore, refer to G. Reiter J. Lindorfer (2013) Möglichkeiten der Integration von Power-to-Gas in das be-stehende Energiesystem. In: Steinmüller, Hauer, Schneider (Hrsg.) Jahrbuch Energiewirtschaft 2013. NWV Verlag.
D7.7 Analysis on future technology options and on techno-economic optimization Page 9 of 89
industries in terms of material utilization. In the steel industry, for instance, hydrogen can be used
as a reducing agent in pig iron production (hydrogen reduces iron ores by removing the containing
oxygen) to aid in low-carbon steel production. Instead of reformers using natural gas to produce
hydrogen, it would be possible to shift to carbon-neutral hydrogen produced in electrolysis plants
under certain conditions (if there are no natural gas pipelines or only low amounts of hydrogen
available at a certain location).
Therefore, endeavors toward the decarbonization of the European energy system must be consid-
ered an opportunity to boost European leadership in innovative energy technology, energy-related
transport technology and services, and in the application and implementation of mature, green gas-
related technologies. European policies often intend the direct usage of electricity. However, this
faces restrictions and limits, which can be effectively negated by transitioning to gaseous green
sources like PtG products, green hydrogen, and green synthetic natural gas (SNG). Although its
technological efficiency is relatively low, the production of SNG allows for the unrestricted use of the
existing natural gas infrastructure and offers a completely mature technology and market availability
for all system-relevant components, from storage to the final consumer.
The main objective of work package WP7 of the STORE&GO project is dealing with the technologi-
cal, economic, regulatory, environmental, and social barriers that must be reduced for PtG to be
successfully implemented. Task 7.2 addresses the techno-economic optimization of the PtG system,
focusing on reducing investment costs through experience curves, learning effects, and economies
of scale. This Deliverable D7.7, “Analysis on future technology options and on techno-economic
optimization,” examines the investment cost reductions enabled by PtG applications through econ-
omies of scale (in this Deliverable, the term “economies of scale” refers to the effect of cost reduction
through upscaling). This Deliverable is based on Deliverable D7.5, “Report on experience curves
and economies of scale,” in which the term “economies of scale” refers solely to the effect of real
cost reductions through increases in production volume, rather to increases in size via upscaling
(e.g., of nominal power). Since the market launch and development of PtG technology depend on,
among other things, the profitability (and thus mainly on the investment costs) of the plant, the po-
tential cost reduction should be examined. In addition to the key technological characteristics (e.g.,
state of the art and future projects), new developments, technologies and materials, and potential
future fields of application are also analyzed. Finally, the SNG production costs are calculated for
different applications in order to demonstrate PtG’s potential. The Deliverable shows that barriers
and prejudices can be reduced to enable the successful implementation of PtG plants.
This Deliverable provides a brief introduction, followed by a short summary of the previous Deliver-
able D7.5 “Report on experience curves and economies of scale,” which serves as a basis for the
calculations of investment cost reductions due to up-scaling (economies of scale). The next chapter
discusses the relevant technological characteristics (state of the art and future perspectives), provid-
ing a theoretical basis for the calculations performed in the economic evaluation. Chapters five and
six analyze new developments, technologies, and materials as well as potential future fields of ap-
plication through comprehensive literature reviews. Finally, an economic evaluation is performed by
calculating the specific SNG production costs for different applications and operating modes. Sensi-
tivity analyses are also performed by varying several key parameters in order to determine the main
drivers of SNG production cost reduction.
D7.7 Analysis on future technology options and on techno-economic optimization Page 10 of 89
2 Investigations on technological learning – Recapitulation of
D7.5 results
This Deliverable D7.7 is based on the investigations of technological learning executed in Delivera-
ble D7.5. Hence, this introductory chapter recapitulates that analysis and its results.
In general, the formal concept of experience curves describe the decline of real costs by a constant
percentage (learning rate) for every cumulative doubling of its produced volume and therefore rep-
resents a relationship between the costs of a product and the experience, expressed in cumulative
production of that product.
Note: In that context, economies of scale have also been investigated, in terms of the effect of real
cost reductions through an increase in production volume but not that of cost reductions through size
increases via upscaling (e.g., increases in nominal power), which are investigated separately in this
Deliverable D7.7.
Cost reductions based on experience curves and economies of scale are due to the following fac-
tors, among others:
fixed cost degression (increased utilization of different sectors in the company, such as ad-
ministration, R&D, production, logistics, and distribution),
reduction of production time (increased manpower efficiency due to learning effects),
increased specialization (standardization, focus on core competence and product family),
variation in resources (e.g., alternative and less expensive (raw-)materials, optimized em-
ployment of staff according to qualifications),
improved production technologies,
optimization of product design to simplify the production process.
The produced volume of PtG plants, and therefore the gained experience and economies of scale
depend on the development of the future global demand for PtG products, which is subject to
climate and policy measures (e.g., carbon taxes, the scope of government R&D, subsidies, and
market introduction programs) and economic factors (e.g., economic growth).
While the literature’s data on learning rates, investment costs, and future global demand for PtG
products would principally allow a preliminary estimation of future investment costs for PtG applica-
tions, the available data do not meet our requirements, as they do not differentiate between different
electrolysis or methanation technologies, between systems, or between stack (electrolysis) or reac-
tor (methanation) costs. To obtain a detailed view of technological learning, a component-based
approach was developed with the CoLLeCT (Component Level Learning Curve Tool) model. This
model allows for comparisons of learning effects between different technologies, the investi-
gation of cost structure developments, and a consideration of spillover effects from concur-
rent technology sectors. The potential for cost reductions through technological learning has been
investigated for electrolysis and methanation systems.
Implementing the theory of learning curves requires estimating global PtG demand. Depending on
the scenario, there would be a need to install about 6,500 to 14,200 GW electrolysis power ca-
pacities and about 3,400 to 7,100 GW SNG-output power capacities to meet the demand in
2050. These values seem to be very high. However, it is important to remember that, in a decarbon-
ized energy system in 2050, not only natural gas but also other fossil energy sources such as oil and
coal must be replaced by renewable energy carriers. Since not all areas of the energy system can
be electrified, green molecules (renewable SNG and hydrogen produced by PtG) are also expected
to play an important role in the future energy system. To cover this relatively high demand and
D7.7 Analysis on future technology options and on techno-economic optimization Page 11 of 89
produce the required quantities (about 285,000 electrolyzer systems with an installed power of
50 MW would be required), mass production would be necessary. However, this implies products
with a standardized and mass production-ready design (e.g., no individual installation planning or
piping). The PtG systems must be planned on the basis of greenfield construction (with an interface
power supply, gas connection for feed-in, and, possibly, a CO2 supply) to meet the requirements of
mass production.
The costs are stated as real costs (reference year 2017, €2017). This means that the inflationary
effects that are anticipated and will lead to rising nominal costs have not been considered. Addition-
ally, no significant changes in technology, such as the implementation of additional functions,
control elements and safety devices, or efficiency improvements, have been taken into account in
the calculation of future investment costs. Solely this approach, of assessing the product according
to the current functional scope and characteristics, allows for the investigation of future costs based
on the theoretical concepts of experience curves and economies of scale.
The results indicate that alkaline electrolyzer (AEC) systems show lower potential for cost reductions
than the proton exchange membrane electrolyzer (PEMEC) and solid oxide electrolyzer (SOEC).
The AEC’s estimated investment costs of about 440 €2017/kWel in 2050 are expected to be signifi-
cantly higher than those for PEMEC systems, expected to be about 290 €2017/kWel. Besides the
AEC’s lower overall learning rate, this result may be due to the substantially higher starting value of
cumulative productions, which means that significant learning effects have already occurred. Addi-
tionally, the PEMEC’s learning rate decreases quickly along with increasing production volumes in
the beginning, whereas this effect decreases at higher cumulative volumes. Conversely, the experi-
ence rate of the AEC is more harmonized over the entire period. The SOEC shows the highest
cost-reduction potential of all three investigated electrolysis technologies, with investment
costs estimated to reach about 530 €2017/kWel in 2050.3 This follows from a rather high learning
rate that was defined on the SOEC itself, based on the relevant literature. Even though, calculations
for the SOEC have been specified in more detail in the meantime, especially for this technology,
further investigations into cost structures and experience rates are still necessary to allow reasona-
ble estimations of future investment costs.
The experience curves for catalytic and biological methanation systems show similar cost-
reduction trends. The investment costs for biological methanation are lower in the long term. This
is mainly driven by the fact that the increase in the cumulative produced volume has to be substan-
tially higher than that for the catalytic application to reach the presumed technology production share
levels. Additionally, biological methanation lacks the catalyst component that catalytic methanation
includes; the latter is expected to obtain learning effects that are low compared to those of other
components in the reactor module. However, the investment costs for both technologies remain
on a similar level throughout the investigated period and are expected to reach values of
280 €2017/kWSNG (catalytic) and 220 €2017/kWSNG (biological), respectively, in 2050 under the pre-
sumed conditions.
However, it has to be pointed out that the development of PtG technology is subject to fundamental
energy and climate policy decisions.
3 The calculations of technological learning for the SOEC have been improved, resulting in lower costs com-pared to the values stated in deliverable D7.5. Hence, the base values for the calculations in this deliverable represent up-to-date results for that technology.
D7.7 Analysis on future technology options and on techno-economic optimization Page 12 of 89
3 Economies of scale
Unless otherwise mentioned, cost predictions for the PtG technology in this Deliverable are stated
as real costs (reference year 2017, €2017). This means that the inflationary effects that are antici-
pated and will lead to rising nominal costs have not been considered. Additionally, no significant
changes in technology, such as an implementation of additional functions, control elements and
safety devices or efficiency improvements, have been taken into account for calculating the future
investment costs.
The term “economies of scale” is used in the literature to describe two different forms of cost reduc-
tion for a product. An EoS that directly affects the production process of a certain technology by
going from unit, to batch, and then to series production, leading to reduced unit cost, is considered
part of technological learning and is therefore included in the STORE&GO Deliverable D7.5 “Report
on experience curves and economies of scale.” This Deliverable analyzes reductions in specific in-
vestment costs for individual PtG plants through the upscaling of nominal power, according to the
reference value. The term “EoS” in this Deliverable refers solely to the effect of cost reduction at-
tained through an increase in size/scale/power via upscaling (e.g., of nominal power).
Using a logarithmic relationship is a common method of estimating costs by scaling. This is known
as the “six-tenth-factor rule” [1], or the “scale factor” or “cost-to-capacity” method:
𝐶𝑏 = 𝐶𝑎 ∗ (𝑆𝑏𝑆𝑎)𝑓
Eq. 1
where 𝐶𝑏 stands for the questioned equipment costs at the appropriate scale 𝑆𝑏 (size, capacity,
nominal power) of the component, and 𝐶𝑎and 𝑆𝑎 represent the costs and scale of the known refer-
ence component, respectively. 𝑓 is the scale factor applied to the technology in question. If no other
information is available, 𝑓 = 0.6 can be used as a scale factor in an initial approximate cost estima-
tion (this is where the term “six-tenth-factor” comes from) [1].
However, the value of the scale factor 𝑓 is specific to the component because the influence of equip-
ment scaling on cost is related to the design and structure of the individual component. PtG systems
consist of a variety of individual components, resulting in a wide range of scaling effects and influ-
ences on the overall system costs. To enable an accurate estimation of the influence of scaling for
PtG systems, these EoS are investigated in detail below. Based on the work on technological learn-
ing in the STORE&GO Deliverable D7.5 [2], a modular approach is taken by splitting up the investi-
gated systems into separate modules and components. Furthermore, electrolysis and methanation
systems are again investigated separately. Splitting the plant into individual modules and using sev-
eral scale factors reduce the risk of using a single inappropriate scale factor for the entire plant.
3.1 Electrolysis
The electrolysis systems analyzed herein – based on AEC, PEMEC and SOEC technology – can or
must be different in design depending on the requirements / framework conditions / operation pur-
pose (e.g. required gas quality and conditions, heat management, and gas drying). This results in a
large number of variants of individual electrolysis concepts, which also differ in investment costs.
Since not all possible variants can be analyzed in this study, the investment costs calculated thus
serve as a guideline for cost estimations of future projects. The actual investment costs for a specific
D7.7 Analysis on future technology options and on techno-economic optimization Page 13 of 89
project, where the respective requirements or framework conditions in the plant design are consid-
ered, have to be analyzed in detail by the manufacturers and may differ from those estimated herein.
3.1.1 Literature review on EoS of electrolysis systems
Using investment cost data taken from the literature review in Deliverable D7.5 and other sources,
the scale factors for AEC and PEMEC are calculated using equation Eq. 1 and are presented in
Table 3-1. The mean scale factor for both technologies is about 0.75; however, the range is wide,
from about 0.51 to 0.96 for AEC and 0.53 to 0.97 for PEMEC. This wide range can be attributed to,
among other things, the wide range (from 0.1 to 100 MW) of the analyzed system scales, because
the scale factor for small-scale electrolyzers (< 5 MW) is lower than that for large-scale ones
(> 5 MW). The mean scale factor for AEC is about 0.69 (< 5 MW) and 0.90 (> 5 MW) on average,
and that for PEMEC is about 0.72 and 0.82, respectively. This means that the positive effect due to
upscaling (EoS) declines for large-scale electrolyzers.
Table 3-1: Calculated scale factors for electrolysis systems based on cost data from literature
Nominal power
Spec. invest-ment costs
Investment costs
Scale factor – rel. to the previous scale
Mean scale factor
Cost data based on source
/MW /€/kW /Mio. € - - -
AEC 0.5 1,800 0.9 -
0,75
[3] 2.5 1,200 3.0 0.75
0.5 2,000 1.0 -
[4] 1.0 1,500 1.5 0.58
10.0 1,000 10.0 0.82
0.4 2,370 0.8 - [5]
3.4 875 2.9 0.56
0.5 1,893 0.9 -
[6] 1.0 1,795 1.8 0.92
2.0 1,746 3.5 0.96
0.7 2,521 1.9 -
[6] 1.5 1,845 2.7 0.54
2.3 1,473 3.4 0.51
1.0 1,150 1.2 -
[7] 5.0 710 3.6 0.70
10.0 682 6.8 0.94
50.0 620 31.0 0.94 PEMEC
0.1 3,500 0.4 -
0,75
[3] 1.0 1,750 1.8 0.70
0.6 2,915 1.7 - [5]
3.0 1,370 4.1 0.53
5.0 1,130 5.7 - [8]
30.0 940 28.2 0.90
0.6 2,250 1.2 -
[6] 1.1 1,715 1.9 0.61
2.2 1,390 3.1 0.70
1.0 1,943 1.9 - [6]
2.0 1,598 3.2 0.72
0.5 1,450 0.7 -
[9]
1.0 1,300 1.3 0.84
2.5 1,050 2.6 0.77
5.0 1,000 5.0 0.93
10.0 750 7.5 0.58
100.0 700 70.0 0.97
D7.7 Analysis on future technology options and on techno-economic optimization Page 14 of 89
A similar scale factor of 0.7 (based on previous studies) is also presumed for use in calculating the
influence of economies of scale, as in [10].
For SOEC systems, the available data (especially on systems with capacities above 500 kW) are
insufficient to allow analysis comparable to what is shown in Table 3-1 for AEC and PEMEC. Though
the components are comparable (at least in terms of production and scaling) and the EoS effects
are thus expected to be in a similar range, the cost structures of SOEC systems are very different.
Therefore, a reliable estimation cannot use data from the literature but requires a more detailed
investigation. Such an analysis is performed below for all three technologies.
3.1.2 Calculation of specific investment costs of electrolyzer systems due to EoS
At a minimum, the following data are needed to estimate the development of the specific investment
costs of electrolysis systems for different nominal power ranges resulting from EoS in a modular
approach:
total investment costs of the plant at a reference scale,
appropriate cost shares of the individual modules, and
their corresponding scale factors.
Table 3-2 presents the specific investment costs for 5 MWel electrolyzer systems (AEC, PEMEC,
and SOEC) in 2020, 2030, 2040, and 2050. The determination of these costs has already been
extensively discussed in STORE&GO Deliverable D7.54 [2]. The cost reduction is based solely on
experience/learning curve effects due to an increase in the cumulative production volume. The costs
shown in Table 3-2 serve as a reference value (5 MWel nominal power) for the calculation of invest-
ment costs for electrolyzers with a nominal power in the range of 1–100 MWel through EoS.
Table 3-2: Calculated specific investment costs for a 5 MW electrolysis systems due to learning curves [2]
Year of installation
Specific investment costs [€/kWel]
AEC PEMEC SOEC
2020 1,060 970 1,990
2030 760 530 1,060
2040 510 340 660
2050 440 290 530
The development of the cost structure of the 5 MWel electrolyzer reference systems until 2050 (see
Figure 3-1) was also calculated according to STORE&GO Deliverable D7.5. Since not all compo-
nents or modules are affected by technological learning to the same extent, the cost structures of
the systems change as cumulative production grows [2].
4 The learning curves, and therefore the resulting costs of the SOEC systems, have changed slightly from the results shown in deliverable D7.5 because the underlying component structure specification has been refined and become more detailed based on more recently available literature data.
D7.7 Analysis on future technology options and on techno-economic optimization Page 15 of 89
Figure 3-1: Development of the cost structure of a 5 MW electrolysis system due to learning curves [2]
A scale factor (see Table 3-3) is defined for each module of the electrolysis system, consisting of
cell stack, power electronics, gas conditioning, and balance of plant (BoP).
Table 3-3: Scale factors of the main parts of an electrolysis system
Component Scale factor
AEC PEMEC SOEC
Stack (initial) 0.88 0.89 0.87
Power electronics 0.75
Gas conditioning 0.60
BoP 0.68 0.73 0.73
The scale factor for the stack module is investigated in terms of the underlying components and is
calculated as the product of the varying cost structure (according to the learning curves) and individ-
ual scaling effects (a detailed compilation is provided in the appendix). Thus, the scale factor varies
with the changing cost structures due to the learning effects (cf. Deliverable 7.5 [2]). The stack does
not show potential for large cost reduction via EoS because of its modular design (cf. [11]). An in-
crease in stack power due to an upscaling of the electrolyzer cell is unlikely for many reasons (e.g.,
problems with leakage); therefore, the cell is limited in size. This maximum cell stack size is expected
to increase as TRL and technological advances increase. To take those effects into account, a dy-
namic scale factor is implemented for the electrolysis cell stack based on an exponential function:
𝑓 = 1 − (1 − 𝑓0) ∙ 𝑒−𝑆𝑆0 Eq. 2
where 𝑓0 represents the basic scale factor as shown in Table 3-3, 𝑆 is the questioned scale, and 𝑆0
is the average maximum stack size for the period under study. This provides a scale factor that is
dependent on the system scale itself and minimizes scaling effects for large-scale applications.
D7.7 Analysis on future technology options and on techno-economic optimization Page 16 of 89
Figure 3-2: Dynamic scale factor for electrolysis cell stack
The following table shows the presumed average maximum stack size for the electrolysis technolo-
gies and installation years.
Table 3-4: Average maximum stack sizes used for electrolysis cell stacks per year of installation, based on [12]
Year of installation
avg. max. stack size S0 [MWel]
AEC PEMEC SOEC
2020 3.0 1.2 0.5
2030 4.0 2.0 1.0
2040 5.0 3.5 2.0
2050 5.0 5.0 3.0
Following [11], a scale factor of 0.75 is used for the power electronics (transformer and rectifier). The
gas conditioning module consists mainly of components for gas drying and cooling (with an average
scale factor of 0.52 [11,13]) and H2 purification (with a factor of 0.81 [11]). Thus, an average scaling
of 0.60 is used. Following data in the literature (cf. [1,11,13,14]) an average scale factor of 0.68 to
0.73, depending on the underlying system and its components, is defined for the overall module,
which is a mixture of many different components, including piping (1.33), heat exchangers (0.59),
valves and fittings (0.60), pumps (0.59) and further equipment. For BoP components, where no scale
factor is available in the literature, an average factor for BoP of 0.6, according to the six-tenths rule,
is assumed. These assumptions are largely supported by the data provided by [1]. A detailed com-
pilation of the incorporated components and the presumed scale factors is presented in the appen-
dix.
Based on the data described above (i.e., specific investment costs of the reference system, cost
structure, and scale factors), the specific investment costs are calculated according to Eq. 1 for each
individual module and subsequently summed up for the entire electrolysis system. The results for
the development of the specific investment costs for electrolysis systems due to EoS for a nominal
power of 1–100 MW in 2020, 2030, 2040, and 2050 are shown in Figure 3-3 for AEC, Figure 3-4 for
PEMEC, and Figure 3-5 for SOEC.
D7.7 Analysis on future technology options and on techno-economic optimization Page 17 of 89
Figure 3-3: Specific investment costs of AEC due to economies of scale for a nominal power of 1-100 MW in 2020,
2030, 2040, and 2050
Figure 3-4: Specific investment costs of PEMEC systems due to economies of scale for a nominal power of 1-100 MW in
2020, 2030, 2040, and 2050
D7.7 Analysis on future technology options and on techno-economic optimization Page 18 of 89
Figure 3-5: Specific investment costs of SOEC systems due to economies of scale for a nominal power of 1-100 MW in
2020, 2030, 2040, and 2050
The average scale factor for the entire electrolysis system can be determined based on the overall
investment costs for the electrolysis system calculated from the individual modules. Figure 3-7 shows
the development of the range and average scale factor for electrolysis systems with a nominal power
of 1–100 MW in 2020, 2030, 2040, and 2050.
The resulting scale factor is affected by two parameters: (1) the system scale itself, due to the mod-
ular approach, which is intensified by the dynamic scaling of the stack; and (2) the year of installation,
due to shifting cost shares as a result of technological learning. These effects can clearly be seen in
Figure 3-6 for the AEC electrolysis systems.
Figure 3-6: Development of module cost shares for AEC electrolysis systems in dependency of the system scale and the
year of installation
The effect of EoS is more pronounced at lower nominal power levels than at higher levels. For ex-
ample, the scale factor for AEC is about 0.78 for 1 MW and 0.88 for 100 MW in 2020 (see Figure
3-7). This peculiarity is due to the changes in the electrolysis system’s cost structure caused by the
D7.7 Analysis on future technology options and on techno-economic optimization Page 19 of 89
different scale factors of the individual modules. Due to the rather high scale factor (up to 1 in large
systems), the costs of the stack do not decrease as much as do, for example, the costs of the gas
conditioning (scale factor = 0.60). Relative to the reference system, this leads to an increased share
of stack costs in the overall system for larger electrolysis systems, which reduces the overall effect
of EoS.
Further, the effect of EoS is stronger in the future than in the present. For example, the scaling factor
for the PEMEC reference system (5 MW) is calculated at 0.87 on average in 2020 but 0.78 in 2050.
This difference is due to the development of the cost structure through the effects of technological
learning (see Figure 3-1). As cumulative production increases, the module costs for the stack de-
crease due to learning effects more steeply than do, for instance, the costs of the gas conditioning
module. For this reason, the scale factor of the entire electrolysis system declines in the future, since
the modules with low learning-curve effects (gas conditioning, power electronics, and BoP) are more
appropriate for scaling and represent higher shares of the overall system costs. Therefore, they
dominate in the cost structure, which leads to a stronger overall effect of EoS.
The calculated scale factors for AEC and PEMEC are in a similar range, of about 0.75 to 0.90, though
the average values are a little higher in the PEMEC case, particularly in early periods. The ranges
for the SOEC system are mostly smaller, at about 0.74-0.82, due to the significantly differing cost
structure and the fact to profit more from scaling due to the lower overall scale factors.
Figure 3-7: Future development of the range and average scale factor for electrolysis systems with a nominal power of
1–100 MW in 2020, 2030, 2040, and 2050
Comparing the calculated scale factors for AEC and PEMEC systems using the modular approach,
with the values calculated from the analyzed literature cost data, shows that the limits of the former
are closer, especially on the lower side. This occurs, first, because small applications below 1 MWel
that show low scale factors are not considered in our analysis. Second, some of the values calculated
based on the literature cost data seem to be rather are low compared to scale factors of individual
components found in the literature and used in the modular approach. This may indicate that the
cost reductions in the analyzed literature data incorporate effects other than EoS (according to the
definition used in this study), such as learning curve effects. Moreover, the values are not compara-
ble, as the values in the literature use varying reference values due to data limitations, while our
calculations use a fixed reference of 5 MWel.
D7.7 Analysis on future technology options and on techno-economic optimization Page 20 of 89
3.2 Methanation
The methanation systems analyzed herein – catalytic and biological – can be further subdivided into
various processes and reactor technologies (e.g. for catalytic methanation reactor: fixed bed, fluid-
ized bed, coated honeycomb, bubble column). The individual concepts can or must be different in
design depending on framework conditions / requirements / operation purpose (e.g. gas qualities
and conditions, reactor concept and stages, heat management, and gas drying). This results in a
large number of variants, which also differ in the investment costs.
Since not all possible variants can be analyzed in this study, the investment costs calculated thus
serve as a guideline for cost estimation of future projects. The actual investment costs for a specific
project, where the adaptations of the methanation plant to the respective framework conditions are
considered, have to be analyzed in detail by the manufacturers and may differ from those estimated
herein.
3.2.1 Literature review on EoS for methanation systems
The scale factors for biological and catalytic methanation systems are calculated using equation Eq.
1 based on the investment cost data taken from the literature review in Deliverable D7.5 (see Table
3-5). The mean scale factor for biological methanation systems is about 0.52 (range 0.39–0.73), and
that for catalytic systems is about 0.64 (range 0.58–0.71). As mentioned, these values are based
solely on cost data taken from the literature, where the investment costs are estimated for further
analyses, since no commercial plants are offered by the manufacturers. Therefore, these values can
be used only as a rough guideline.
Table 3-5: Calculated scale factors for methanation systems based on cost date from literature
Nominal power
Spec. invest-ment costs
Investment costs
Scale factor – related to the previous scale
Mean scale factor
Cost data based on source
[MW] [€/kW] [Mio. €] - - -
Biological 0.2 320 0.06 -
0,52
[15] 1 120 0.12 0.39
2 90 0.18 0.58
1 1,439 1.44 -
[16] 10 371 3.71 0.41
20 243 4.86 0.39
50 168 8.38 0.60
1 1,200 1.20 - [17]
110 340 37.40 0.73 Catalytic
1 1,500 1.50 -
0,64
[15] 3 1,000 3.00 0.63
6 750 4.50 0.58
5 300 1.50 -
[18] 30 160 4.80 0.65
110 110 12.10 0.71
5 400 2.00 - [19]
110 130 14.30 0.64
Ref. [20] analyzed the cost improvement in chemical process technologies and identified a scaling
factor of 0.56, based on 20 processes. While some are comparable to methanation, like Lurgi gasi-
fication or ammonia and ethylene production, others are substantially different, like the direct reduc-
tion of iron ore or oil sands extraction.
D7.7 Analysis on future technology options and on techno-economic optimization Page 21 of 89
Ref. [21] reports that recent biorefinery installations exhibited a scaling factor on capital costs in the
range of 0.63 to 0.72.
3.2.2 Calculation of specific investment costs of methanation systems due to EoS
This section calculates the specific investment costs for methanation systems with a nominal SNG
output power in the range of 1–100 MW for 2020, 2030, 2040, and 2050 using the scale factor
method (see Eq. 1) on a modular basis. The analysis does not use a single scale factor but divides
the entire methanation system into individual modules, each with a separate assigned factor. Using
this measure increases the accuracy of the cost estimation because the individual components react
separately to the EoS effect.
The specific investment costs shown in Table 3-6 are used as initial values for the cost estimation
by scaling for 5 MWSNG methanation systems in 2020, 2030, 2040, and 2050. These investment
costs are calculated in STORE&GO Deliverable D7.5 [2] and are solely based on the experi-
ence/learning curve effects of a 5 MWSNG methanation system due to an increase in cumulative pro-
duction volumes.
Table 3-6: Calculated specific investment costs for 5 MWSNG methanation systems due to learning curves in 2020, 2030,
2040, and 2050 [2]
Year of installation Specific investment costs [€/kWSNG]
Catalytic Biological*
2020 580 600
2030 440 390
2040 320 280
2050 280 240
*Based on expert interviews and current data, the specific investment costs have changed slightly
from those calculated in Deliverable D7.5.
To increase the accuracy of the cost estimation, the methanation system is split into sub-modules
(see Figure 3-8). This breakdown and the development of the appropriate cost shares are shown in
the analysis of learning curve effects in STORE&GO Deliverable D7.5 and are here used to analyze
the effect of EoS. The cost structure change is a result of the modularized determination of learning
curve effects, whereby the individual modules are affected to different degrees.
Figure 3-8: Development of the cost structure of 5 MWSNG methanation systems due to learning curves in 2020, 2030,
2040, and 2050 [2]
D7.7 Analysis on future technology options and on techno-economic optimization Page 22 of 89
For each module defined in Figure 3-8, an appropriate scale factor is assigned, as shown in Table
3-7. As the reactor module represents the primary difference between the two investigated methana-
tion technologies, it is again treated in a more detailed way.
Table 3-7: Scale factors of the modules of methanation systems
Component Scale factor
Catalytic Biological
Reactor (initial) 0.67 0.51
Reactor 0.56 0.50
Catalyst 1.00 -
Heat Management 0.56
Electric installation 0.75
Gas conditioning 0.60
BoP 0.67
As was done for the stack modules of electrolysis systems, the scale factor for the methanation
reactor module is calculated, based on the underlying components, as the product of cost shares
and individual scaling exponents. Since the cost structure varies due to learning effects according to
the installation time (cf. Deliverable 7.5 [2]), the resulting overall module scale factor is expected to
be nonconstant as well.
The reactor is different for both technologies: for the catalytic reactor, a scale factor of 0.56 is used
according to [1]; the biological case presumes an agitated reactor with an appropriate scale factor of
0.50 (cf. [13]). The catalyst, which represents the catalyst material itself in the catalytic methanation
reactor, is not subject to EoS, as catalyst usage is presumed to be directly proportional to the nominal
power and thus has a scale factor of 1. For heat management, mainly heat exchangers, a scale
factor of 0.56 is assumed (cf. average values for heat exchangers in [1,13]). Additionally, especially
the reactor for biological methanation does not show potential for large cost reduction via EoS be-
cause it is limited in design size for reasons like plant construction and transportation due to the
height and diameter of the reactor. To increase the power of the SNG plant a numbering-up of reac-
tors is necessary. To consider those effects, a dynamic scale factor is implemented for the reactor
(cf. Eq.2), to provide a scale factor that is dependent on the system scale itself and minimizes scaling
effects for large-scale applications.
The following table shows the presumed average maximum reactor size for biological methanation
plants and installation years.
Table 3-8: Average maximum reactor sizes used for biological methanation plants per year of installation
Year of installation
avg. max. reactor size S0 [MWSNG]
biological
2020 2
2030 5
2040 5
2050 5
D7.7 Analysis on future technology options and on techno-economic optimization Page 23 of 89
Based on the power electronics of the electrolyzer, a scale factor of 0.75 is chosen for the electrical
installation module of the methanation system. Based on data in the literature on gas conditioning
components, mainly consisting of equipment for drying and cooling (scale factor: 0.52) and SNG
purification (0.81), an average scale factor of 0.60 is set for that module. The BoP module includes
many different components, like pumps, valves, tanks, fitting, piping, sensors, frame, and housing.
According to the six-tenths rule, a scale factor of 0.60 is set for components for which data are
missing (this assumption is supported by data in [1] and represents an average value). In combina-
tion with the scale factors found for the other components of the module (cf. [1,13]), an average
medium scale factor for BoP of 0.67 is assumed. A more detailed compilation of the values and
components used is provided in the appendix.
The investment cost of methanation plants in a power range of 1–100 MW are calculated for 2020,
2030, 2040, and 2050 by applying equation Eq. 1 to estimate the investment costs by scaling using
the above data (specific investment costs for the 5 MW reference system, cost structure, and scale
factors for the modules). The results are shown in Figure 3-9 and Figure 3-10 for catalytic and bio-
logical methanation systems, respectively.
Figure 3-9: Specific investment costs of catalytic methanation systems due to economies of scale for a nominal power of
1–100 MWSNG in 2020, 2030, 2040, and 2050
D7.7 Analysis on future technology options and on techno-economic optimization Page 24 of 89
Figure 3-10: Specific investment costs of biological methanation systems due to economies of scale for a nominal power
of 1–100 MW in 2020, 2030, 2040, and 2050
A single scale factor for the entire methanation system can be derived from the investment costs
calculated using the modular approach, as shown in Figure 3-11. The calculated average scale fac-
tors for both methanation technologies are in the same range, of about 0.68 to 0.73 depending to
installation time, since the cost structures do not significantly change. The difference in scale factor
between smaller (1 MW) and larger (100 MW) systems range from 0.02 to 0.09. Larger systems
have a higher scale factor due to the necessity of numbering up the reactors, since a larger reactor
is not possible for manufacturing reasons (for biological methanation an average maximum reactor
size of 5 MW is assumed).
Figure 3-11: Future development of the range and average scale factors for methanation systems with a nominal SNG
output of 1–100 MW
The results for the catalytic methanation system are close to the trends identified in the literature.
The average values for the biological system calculated via the modular approach are a little higher
than those in the literature. This is partly because this Deliverable’s modular approach does not
consider smaller-scale systems with capacities below 1 MW, given our focus on large-scale storage
systems. Nevertheless, the investigations above suggest that scale factors < 0.60 seem to be rather
D7.7 Analysis on future technology options and on techno-economic optimization Page 25 of 89
low. Therefore, it must be assumed that additional effects, like technological learning, are incorpo-
rated in the literature’s data and that the values shown in Table 3-5 do not consider only EoS.
3.3 Alternative to scaling up: Modular design/Numbering up
Besides economies of scale, numbering up due to a modular plant design is another way to increase
the nominal power of a PtG plant. A modular PtG plant structure can offer economic and technical
advantages. The economic advantages include the following [22]:
Costs: lower production costs due to series components;
Delivery date: delivery time is shortened and on-time delivery increases massively
Production time: low cycle time for modular systems (faster configuration and commissioning)
Quality: larger quantities require and allow more time for development and design (higher
cost pays off due to economies of scale)
Flexibility: modules can be changed, replaced, modernized; thus, machines become more
flexible and versatile, and their service life increases
Service: better service options and faster reparation through standardized modules
Operating factor: clear operating communication of modular machines leads to simple and
intuitive operation
All the cost reductions due to series components, the reduced production time, and the increased
quality lead to a reduction in investment costs. These were considered in the analyzes of learning
curves in STORE&GO Deliverable D7.5 [2]. The modular design can also offer advantages in terms
of technical parameters like plant efficiency, H2 and SNG output, and lifetime, which in turn influence
the economy of the plant. However, a modular design also has the disadvantage of losing the effect
of EoS (especially in the case for methanation units).
In the modular design of a PtG plant, a distinction must be made between the two main compo-
nents—the electrolyzer and the methanation—as these have fundamentally different structures. An
electrolyzer is built up of cells, coupled to form a stack, and then connected to build an electrolyzer
module; if necessary, they are clustered for large capacities. Therefore, a part of each electrolyzer
has a modular design. Methanation, which is based on the principle of classical plant engineering,
behaves differently.
The modular design of electrolyzers can be divided into two levels. On the micro level, a large num-
ber of identical cells are stacked one on top of the other to build the electrolyzer cell stack (see Figure
3-12).
Figure 3-12: PEM electrolyzer stack [23]
On the macro level, a number of identical electrolyzer stacks are connected to an electrolyzer system
to increase the rated output. As mentioned, the STORE&GO project “Innovative large-scale energy
D7.7 Analysis on future technology options and on techno-economic optimization Page 26 of 89
storage technologies and PtG concepts after optimization” deals with large-scale PtG concepts. Cov-
ering global PtG demand in 2050 would require the installation of about 6,500 to 14,200 GW elec-
trolysis power capacities (see estimation on PtG demand in Deliverable D7.5). Not only a large num-
ber of electrolyzers are necessary, but they must also have a correspondingly high rated power.
Even the manufacturers of electrolyzers have recognized the need for plants with a high nominal
output and are offering even already today, systems up to the three-digit MW range. So far, however,
no plants of this size have been realized. The largest electrolyzers are used in such projects as
H2Future (with 6 MW) and REFHYNE (with 10 MW). Due to the technical limitations (the nominal
power of a single stack is limited due to design features such as the sealing of the cells), these high
rated outputs can be achieved only with a modular design. Figure 3-13 shows examples of a modular
electrolyzer design on a macro level.
To achieve high rated output (> 10 MW), several stacks/electrolysis modules must be interconnected
(see Figure 3-13), since a stack, as well as the resulting electrolysis module, is limited in size (nom-
inal power). Currently, a single electrolyzer stack has a nominal power of about 2 MW (cf. [24], [25],
[26]). Therefore, a modular design is mandatory. However, the number and size of electrolysis mod-
ules/stacks are not selectable or changeable but are determined by the manufacturer. Module con-
trol and operation are ideally optimized. This also includes the steady state and transient behavior.
For small electrolyzers, with a capacity < 2 MW, it is possible to choose between an electrolyzer
consisting of a single stack or one consisting of several smaller stacks (i.e., modular design).
7 Current density A/cm² 0,5 0,7 0,8 2,0 2,2 2,5 - - -
8 Use of critical raw material as catalysts
mg/W Co 7,3 3,4 0,7 - - - - - -
mg/W PGM - - - 5,0 2,7 0,4 - - -
mg/W Pt - - - 1,0 0,7 0,1 - - -
KPIs and values based on [35] 1) Degradation is defined differently for high and low temperature electrolysis: AEC/PEMEC: percentage efficiency loss when run at nominal capacity; SOEC: percent loss of production rate at thermo-neutral conditions and constant efficiency
D7.7 Analysis on future technology options and on techno-economic optimization Page 32 of 89
4.1.2.1 Operational characteristics
The three technologies under investigation—AEC, PEMEC, and SOEC—differ significantly in their
operational characteristics only in terms of their principles of operation. AEC and PEMEC represent
low-temperature electrolysis, both supplied with liquid water and conventionally operated at temper-
atures of 80 to 90 °C. Operation at higher temperatures would be preferable by means of electric
energy consumption, which is driven by reversible cell voltage. In the conventional case, alkaline
cells are limited to between 100 and 120 °C using commercial diaphragms [36,37], while the Nafion®
membranes used in PEM cells are known to loose water, and thus ionic conductivity, at temperatures
above 100 °C [38]. Though laboratory applications are tested at elevated temperatures of 200 °C
and beyond using alternative materials and solvents (cf. [36] and [38]), significant research efforts
have to be made in material science to find a way to outperform commercial cells, including in eco-
nomic terms. Therefore, an increase in operational temperatures for alkaline and PEM beyond
100 °C is not expected in the foreseeable future [33–35,37].
For solid oxide electrolysis, operating temperatures are typically in ranges of 650 to 1,000 °C [39].
On the lower side, this range is limited to about 600 °C, which allows a sufficiently quick start-up
from standby in transient operation [33]. Generally, electric energy demand decreases as cell tem-
perature increases, while the share of reaction enthalpy that can be covered by thermal energy
increases. Hence, high-temperature electrolysis is more beneficial whenever an external heat supply
is accessible.
Regarding storage densities, the compression of gaseous hydrogen is an energy-demanding task
that significantly decreases overall system efficiencies. The elevation of operating pressure for elec-
trolysis is a potential way to provide high storage densities. However, pressurized operation depends
largely on the manufacturer’s design choice and system philosophy. Systems with output pressures
of up to 80 bar are reasonable in the near future, depending on the demand, thus eliminating the
first stage of external compression and allowing direct feed into distribution gas grids [12,40]. Though
pressurized systems of 100 bar and above for the direct usage at hydrogen fuel stations have been
investigated by several manufacturers and research projects, no widespread rollout is expected
within the next few years [41].
Concerning current electrolysis cell densities, the state of the art has not significantly changed in
recent years (cf. Table 4-1). Most of the literature foresees only marginal increases in the intermedi-
ate term, reaching values of 0.8 to 1.0 A/cm² (AEC), 2.5 to 3.0 A/cm² (PEMEC), and about 1.0 A/cm²
(SOEC), respectively (cf. [12,33,40]). Significant increases are expected over the long-term. This is
particularly interesting in the case of alkaline cells, with values of up to 2.0 A/cm² [40], as this could
result in a decreasing CAPEX, which would be a competitive advantage for this already mature
technology. Moreover, PEMEC and SOEC are expected to increase their densities up to a factor of
2 until 2050 (PEMEC: ~3.5 A/cm², SOEC: < 2.0 A/cm²; according to [40]).
4.1.2.2 Capacity
The hydrogen production rate per stack, or, rather, nominal stack capacity, is rapidly rising for all
three electrolysis technologies and is already outperforming expectations, as shown in Table 4-1.
Hence, estimations of the future development of stack and system sizes based on the literature are
difficult. The comprehensive study of E4tech done in 2014 [12] proposed stack capacities of up to
7.8 MWel for alkaline and up to 10 MWel for PEM cells for the intermediate term (2030), which is, at
least for AEC, not that far beyond actual values. Total system capacities are somewhat different; the
estimations for alkaline electrolysis are close to the maximum stack values (tending single stack
systems), while PEM-based processes are expected to use multiple stacks with total capacities of
up to 90 MWel by 2030 [12].
D7.7 Analysis on future technology options and on techno-economic optimization Page 33 of 89
Regarding the active cell area, which partly correlates with the maximum stack capacity, the value
is expected to increase as the technology develops, especially for PEM and solid oxide cells. Future
cell areas may be in the range of
less than 10 m² for alkaline electrolysis,
less than 1 m² for PEM and
less than 0.1 m² for solid oxide electrolysis,
resulting in a difference of one magnitude between each technology [40].
4.1.2.3 Efficiency
At the current state of the art, the electric efficiency of PEM electrolysis is slightly below that of
alkaline technology [34]. This gap may grow in the near future but will narrow over the long term.
According to recent studies, values above 70%el,LHV (related to LHV) on the system level will be
reached [39,40]. The electric efficiency of high-temperature electrolysis is already exceed-
ing 76%el,LHV on the system level [34], and only marginal improvements are expected [40]. Due to
the possibility of heat being supplied from external sources (e.g., industrial waste heat, solar or ge-
othermal energy), the technology already provides electric efficiencies of 100 %el,LHV and above (for
endothermal operation) on the stack level and is primarily a matter of thermal management and the
availability of external heat.
4.1.2.4 Durability
As mentioned, improvements in the durability of electrolysis stacks and systems have been signifi-
cant and have even outperformed the proposed targets in recent years (cf. Table 4-1). An additional
increase in stack lifetimes is expected for all available technologies, reaching values of 125.000 h
and above for low-temperature electrolysis cell stacks (AEC, PEMEC) and up to 100.000 h for SOEC
stacks [40].
On the system level, lifetimes are already at quite competitive levels of 20 to 30 years for alkaline
and PEM technology. SOEC technology is expected to reach a similar lifetime, of about 20 years,
as soon as it is available on the required scale, though little improvement is expected in the longer
term. AEC and PEMEC systems may be able to improve their lifetimes slightly, with alkaline reaching
up to 40 years for large-scale applications [40].
Degradation rates, and therefore stack lifetimes, are highly dependent on operational parameters,
such as operating temperature and pressure. Transient operation also has a significant effect. As
the impacts of start–stop cycles are not yet well-quantified, systematic studies are necessary to un-
derstand these degradation mechanisms [34].
4.1.2.5 Flexibility
As illustrated in Table 4-1, the part load behavior of state-of-the-art electrolyzers is already quite
close to technical limitations. For alkaline electrolysis, the minimum load is limited by the diffusion of
hydrogen across the diaphragm to the oxygen side, which results in flammable mixtures at low pro-
duction rates [34]. Nevertheless, further improvements in AEC technology are expected to reduce
minimum load values to 10% [40].
Limitations in the part load behavior of water electrolysis on the stack level are becoming less im-
portant, as weaknesses in this area can be overcome by using a modular construction for electrolysis
systems, including multiple stacks. In this way, load ranges can be expanded through the adapted
operation of individual units [40].
D7.7 Analysis on future technology options and on techno-economic optimization Page 34 of 89
Start-up times from standby and cold state are also important aspects of operational flexibility. While
it is expected that start-up times from standby will reach similar levels for all three technologies in
the long run at levels below 1 minute, start-up from cold state is far more dependent on the technol-
ogy. Whereas PEMEC systems are already showing reaction times in the range of single seconds,
little improvement is expected for alkaline cells. Major improvements are proposed for high-temper-
ature electrolysis, which is showing long warm-up times of multiple hours. This value may reach the
level of alkaline cells, or even below, in the long run (by 2050), given proper heat management
[34,40].
4.1.2.6 Economic characteristics
Capital costs for electrolysis are expected to significantly decrease, as it gains substantial market
uptake due to learning curve effects and scaling of production, as discussed in STORE&GO Deliv-
erable D7.5 [2]. Additional reductions expected from system scaling are discussed in section 3.1;
these are mainly dependent on future system sizes.
Average operating expenditures are similar for PEMEC and AEC at 2% of CAPEX per year and at
5% of CAPEX per year for solid oxide cells. These values are not expected to change significantly
in the intermediate future, as shown by the data given by FCH 2 JU [35] presented in Table 4-2.
Furthermore, these costs are very sensitive to location and size [12]. Therefore, they have to be
estimated in relation to the individual application.
4.2 Methanation
In this Deliverable, the methanation process covers the production of SNG (synthetic natural gas)
from hydrogen gas, produced in an upstream electrolysis process, and carbon dioxide. This defini-
tion covers very different kinds of process chains, with a primary distinction between biological and
chemical (catalytic) methanation. The KPIs used to evaluate these processes are defined below and
generalize the characteristics of methanation. These KPIs are meant to be used to monitor and
assess the development of the methanation process as part of the PtG concept and provide compa-
rability among the different technologies.
KPIs:
1. Operational characteristics
Process temperature
Operating pressure
GHSV
2. Capacity
3. Conversion efficiency
4. Durability
Catalyst lifetime
Availability
5. Flexibility
Response characteristics
Cold start-up time
6. Economic characteristics
CAPEX
OPEX
D7.7 Analysis on future technology options and on techno-economic optimization Page 35 of 89
4.2.1 SoA and medium-term prospects
Unlike for water electrolysis, few comprehensive reviews have evaluated recent developments in the
underlying processes and applications of methanation technology, likely due to the lack of focus on
that conversion step. In the PtG (or, rather, power-to-methane) process chain, methanation is pri-
marily used to be able to integrate the produced gas into the existing infrastructure, such as for
feeding into regional gas grids. In this context, the additional methanation step is taken into account
in order to facilitate the acceptance of PtG and ensure rapid implementation. On the other hand, the
underlying processes are already quite mature. Chemical CO and CO2 methanation processes have
been investigated for more than 100 years since being discovered by Sabatier and Senderens [42],
while biological methanation is principally comparable to the processes used in biogas plants and
can even be integrated into the same reactor (i.e., in-situ process) [43]. Therefore, most recent de-
velopments have tackled the optimization of reactor technologies, upsizing, and cost reductions
[15,17,42,43].
Table 4-3 gives an overview of the technology characteristics of chemical and biological methana-
tion.
Table 4-3: Technology characteristics for state-of-the-art methanation processes
Parameter Chemical (catalytic)
methanation Biological methanation
Operation
Process temperature (°C) 200-700 [15,19,42] 151)-982) [15,43]
and usually has to be optimized to specific boundary conditions and integrated processes (e.g., HT-
electrolysis, biogas plant). Furthermore, in many applications, the dedicated use-case defines the
mandatory conversion rate and is therefore a determinant of what is necessary for recirculation and
subsequent gas conditioning. If the product gas is to be fed-in to existing gas grids, the related re-
quirements concerning gas quality, especially H2 and CO2 contents, have to be met. Hence, certain
synthetic gas applications do not need the highest conversion rates in the methanation process.
D7.7 Analysis on future technology options and on techno-economic optimization Page 37 of 89
4.2.1.4 Durability
Sensitivity to sulfur compounds in the feed gas is a main cause of catalyst poisoning in chemical
methanation. Therefore, extensive knowledge of gas composition, particularly of CO2 streams from
industrial or biogenic processes, as well as of impurities is mandatory to ensure appropriate catalyst
lifetimes. To achieve lifetimes above several hundred hours, state-of-the-art nickel catalysts require
sulfur contents in the range of several parts per billion [42].
As mentioned, temperatures above 500 °C lead to the thermal degradation of conventional catalyst
materials. Therefore, the application of high-temperature methanation will require the implementa-
tion of alternative catalysts. Appropriate materials are already available on the market but are not
yet widely used [42].
4.2.1.5 Flexibility
While biological methanation reactors are not appropriate for applications with the highest production
capacities, they provide significant advantages in terms of cold start-up time and response charac-
teristic. Therefore, applications of the technology are expected to be used to balance power for fluc-
tuating renewable energies in future energy systems.
In catalytic systems, transient operation often leads to vapor-solid reactions, thermal degradation, or
the crushing of the catalyst [42]. Thus, chemical methanation faces problems that need to be solved
if it is to reach the high flexibility levels attained by its biological counterpart.
4.2.1.6 Economic characteristics
The future development of investment costs for methanation plants have been discussed in this
Deliverable (cf. section 3.2) and STORE&GO Deliverable D7.5 (cf. [2]).
Operational and maintenance costs are expected to be around 10% of CAPEX per year for both
technologies. These include costs for catalyst replacement in the catalytic case and for heating de-
mand and miscellaneous in the biological case [15]. In general, these costs are difficult to estimate
using the literature, as they heavily depend on individual system boundaries and other aspects.
However, the values are expected to be in the upper region and are expected to decrease as de-
ployment increases. Other studies also anticipate such values (cf. [17,50]).
D7.7 Analysis on future technology options and on techno-economic optimization Page 38 of 89
5 New developments, technologies, and materials
Much research has been performed on the implementation of PtG systems as energy storage for
future energy systems. Gruber et al. recently stated that current PtG efficiencies could reach 76%,
and increase to 80% in the future [51]. This section reviews the promising new developments and
findings concerning electrolysis, methanation, and CO2 capturing/separation, the three key technol-
ogies of PtG systems.
5.1 Electrolysis
Research on electrolysis is not focused only on enhancing existing systems (e.g., PEMEC, AEC,
and SOEC); it is also focused on developing new types of electrolyzers. The key factors for improving
electrolysis in the research are efficiency and costs. Research on both factors seeks to enable high-
capacity implementation with comparably low levelized costs of energy.
Ogawa et al. show that research on water electrolysis has been increasing since 1980. Their yearly
publication analysis has shown that the category “Anode and acid stable cathode for AEC and
PEMEC” has the highest publication rate, followed by research on microbial electrolyzer cells (MEC),
SOEC, AEC, and PEMEC. The first two categories of research have increased particularly rapidly
[52]. Gruber et al. state that electrolyzer efficiencies can already reach conversion rates exceeding
90% and efficiency of around 80% [51].
5.1.1 Conventional technologies
Research on established technologies focuses on enhancing the technology by improving individual
parameters. Since higher temperatures would have a positive effect on the reaction kinetics of the
redox reaction as well as on water and heat management, improved thermal stability could increase
the efficiencies of electrolyzers. The main problem with high temperatures in PEMEC is the degra-
dation of the catalyst (e.g., due to carbon corrosion from the usage of a carbon matrix to stabilize
Pt). Specifically, carbon corrosion could cause the Pt catalyst to detach, leading to a severe loss of
platinum surface area [53].
To solve this problem, Pt nanoparticles are produced and stabilized with a siloxane matrix. Pyrolysis
experiments have shown a homogeneous distribution at a temperature of 500°C. Experiments using
a pyrolysis temperature of 600°C have led to Pt agglomeration. The siloxane matrix has shown
slightly higher catalytic activity in the electrolysis process than the carbon-based matrix has. Thus,
siloxane may be a good option for use as a Pt catalyst stabilizer for PEMECs and could increase
efficiency through higher operation temperatures [53].
Tymoczko et al. found that selective Cu positioning could optimize the platinum electrodes used in
PEMEC. Research has shown that monolayer copper has a positive effect on electrolyzer perfor-
mance. The submonolayer of Cu atoms in the second atomic layer of the Pt(111) has shown the
most active electrocatalytic behavior for the hydrogen evolution reaction in acidic media ever re-
ported under comparable conditions [54].
Since most systems need a fresh water supply, operation in regions without sufficient resources
would not be possible. Research is being conducted on the possibility of building electrolyzer sys-
tems that can be operated with seawater. The main problem is that conventional anodes would emit
toxic chlorine and are not sufficiently resistant against degradation [55,56]. One of the most promis-
ing ways is to use molybdenum as electrode material. Fujimura described the possibility of using
Mn-Mo oxide electrodes prepared by anodic deposition on IrO2-coated substrate [56]. It has been
D7.7 Analysis on future technology options and on techno-economic optimization Page 39 of 89
shown that such an electrode has a high oxygen evolution throughout a long period. It has also been
found that Mn-Mo oxides have to be anodically deposited at 90°C to avoid the dissolution of the
oxide. The long-term oxygen evolution efficiency could reach 99.6% with this technology [57]. It has
further been shown that NiFe-layered double hydroxide and Pt nanoparticles are good catalysts for
the electrodes. Dresp et al. used such electrodes for a system that was run only during daytime,
since it used the electricity generated from photovoltaics. SEM investigations have shown a mem-
brane-induced stability loss. During daytime, however, recovering effects have been found. Conse-
quently, such a day–night cycle could be a way to use seawater with conventional electrolysis, by
employing excess electricity drawn from photovoltaic electricity production [55]. Another way of
avoiding the need for a fresh water supply is using high-temperature SOEC. Lim et al. have shown
that contaminants such as sea salt are not found in the steam produced from seawater. Electrolysis
has shown almost the same performance with seawater as was shown with fresh water, as well as
similar degradation levels. The influence on performance of direct electrode contamination with sea
salt was not investigated [58].
Besides the optimization of working temperature and fresh water supply, much of the research fo-
cused on cost reduction has sought new materials for existing electrodes, because current technol-
ogies use expensive noble metals of the Pt group. Reducing or eliminating noble metal content could
reduce costs and make it easier to establish large-scale industrial electrolyzers.
Gabler et al. performed experiments with ultrashort pulse laser-structured nickel electrodes for AEC.
It was found that this technique reduces overpotential by increasing the specific surface area of the
Ni electrode. This method can be improved by activating the Ni electrode with a cyclic voltammetric
reduction-oxidation pretreatment [59]. Hinnemann et al. showed another possibility for new electrode
materials, wherein MoS2 nanoparticles supported on graphite were used as a new type of electrode.
This was shown to be a good alternative to Pt-group metals. Furthermore, Hinnemann et al. claimed
that searching for new electrode materials with a quantum chemical method would be another option
[60]. The catalytic activity of MoSx on hydrogen electrolysis has also been shown by Zeng, et al. [61].
Liu et al. prepared nanohybrids consisting of carbon nanotubes with CoP nanocrystals through the
low phosphidation of Co3O4 nanocrystals. These nanohybrids were shown to be a good electrocat-
alyst for the hydrogen evolution reaction. Therefore, it might be a good option for use as anode
material, especially since it is inexpensive, acid stable, and highly active [62]. According to Jin et al.,
cobalt-cobalt oxide/N-doped carbon sheets hybrids increase the catalytic effect on the hydrogen and
oxygen evolution reaction, and also has high stability as an electrode [63].
Another promising material is nickel-cobalt-iron layered double hydroxide, which has shown excel-
lent electrochemical properties. It is an active electrode material, employed as a positive electrode
with activated carbon employed as a negative one; it is also a good catalyst for the oxygen evolution
reaction. In addition, it has shown good specific power and cycle life. Its positive properties can be
attributed to the synergistic effects among the metal species, as well as to the mesoporous structure
of the layered double hydroxide [64].
Mangan cluster research has shown that graphitic carbon-based electrodes coated with MnOx have
a catalytic effect on the oxygen evolution reaction when used as anodes. It has been found that they
show good stability against corrosion, in contrast to electrodes based on buckypaper and carbon
[65].
The CELL3DITOR project has conducted research seeking an additive manufacturing method for
producing SOEC stacks. It has been reported that it is possible to sinter 8YSZ (yttria-stabilized zir-
conia) particles at temperatures of around 1300°C to reach a relative product density of up to 96%.
This could allow the 3D printing of SOEC stacks in the near future [66,67].
D7.7 Analysis on future technology options and on techno-economic optimization Page 40 of 89
5.1.2 Alternative technologies
Currently, AEC and PEMEC are the main low-temperature electrolyzers for pilot and industrial ap-
plications. Both use a membrane or diaphragm to separate the O2 and H2, and serve as transport
medium for ions between the electrodes (see Figure 5-1 A and B). Esposito describes membraneless
electrolyzers as a low-cost alternative to hydrogen production [68]. In this membraneless case, the
separation of O2 and H2 is performed using fluid flow and/or buoyancy forces. Flow-by electrolyzers
(see Figure 5-1 C) use the laminar flow of the electrolyte, which is parallel to the electrodes, while
flow-through/buoyancy electrolyzers use a pressurized environment to force the electrolyte through
the electrodes into different chambers (see Figure 5-1 D). In both cases, the products are separated.
With this technology, a product purity of up to 99.8% H2 can be achieved [68].
Figure 5-1: Schematic of classic (A & B) and alternative membraneless (C & D) electrolyser cells [68]
The design of membraneless electrolyzers is very simple compared to that of conventional ones,
leading to a long service life, high tolerance to impurities, and operability in extreme conditions. Fur-
ther, it works without an expensive membrane. Both lead to lower CAPEX and allow easier manu-
facturing. Additive manufacturing such as 3D printing could lead to lower investment costs. However,
the high ohmic resistance of the electrolytic solution and the low voltage efficiency at high operating
current densities are the main disadvantages of this technology. Lower purity compared to PEMEC
is also a problem, which may make downstream purification necessary. Future challenges could be
the scaling-up and material-related issues [68].
Experiments with plasma electrolysis have been performed on a small scale. This is a combination
of electrolysis and pyrolysis. The redox reaction occurs without contact between the electrodes and
the water. Electrolysis is performed at temperatures of 3,700°C. In plasma electrolysis, the power
efficiency reaches 30% of the input voltage. This could provide a rapid and cost-effective option in
the near future. Future research will focus on large-scale options for this kind of electrolyzer technol-
ogy [69,70].
D7.7 Analysis on future technology options and on techno-economic optimization Page 41 of 89
Figure 5-2: Schematic of a plasma electrolyzer [70]
5.2 Methanation
As mentioned, two different types of methanation can be used for PtG applications: catalytic and
biological methanation. This section discusses new developments regarding both types.
Most of the research on catalytic methanation focuses on improving the cost and efficiency of the
catalysts. Since the methanation reaction prefers low temperatures (<400°C) due to its exothermic
character, catalysts have to be found that fulfill the necessary catalytic activity, even for low temper-
atures [71,72]. This is required to reach high-quality SNG, since natural gas networks are covered
by strict regulations regarding it. For example, the requirement that must be met for the SNG to be
fed into the natural gas grid is a CH4 content of 96 vol.% or higher [73]. The second largest research
area focuses on gas cleaning (specifically H2 and CO2 separation).
Figure 5-3: Equilibrium for different reactions that occur during methanation. R1 (𝑪𝑶 + 𝟑𝑯𝟐 ↔ 𝑯𝟐𝑶 + 𝑪𝑯𝟒), R2 (𝑪𝑶𝟐 +
𝟒𝑯𝟐 ↔ 𝟐𝑯𝟐𝑶+ 𝑪𝑯𝟒) and R4 (𝟐𝑪𝑶 ↔ 𝑪+ 𝑪𝑶𝟐) are the most important ones during methanation [71]
Studies have shown that the order of catalytic active materials for methanation is
Ru > Fe > Ni > Co > Rh > Pd > Pt > Ir (see Figure 5-4). However, Ru and Co are very expensive
compared to Ni. Therefore, Ni is the most frequently used catalyst. The problem with Ni is that it can
form Ni(CO)4, which has to be avoided since it is highly poisonous at low temperatures. The metals
D7.7 Analysis on future technology options and on techno-economic optimization Page 42 of 89
are usually dispersed on supports as nanoparticles to achieve a high specific surface area for the
catalyst. Al2O3 is one of the most common supports, but SiO2,CeO2, TiO2, ZrO2, and other oxides
are also used [71]. Liu et al. analyzed Al2O3, CeO2, and ZrO2 as support materials and found that
Al2O3 had the highest conversion rate. The second fastest reaction rate was with ZrO2, and the third
was with CeO2. This effect can be seen in Figure 5-5 [74].
Figure 5-4: Activity of different catalysts for the methanation process [71]
Figure 5-5: CO conversion rate on different support materials [74]
Biegger et al. developed an innovative methanation system that uses a washcoated honeycomb
catalyst combined with a polyimide membrane for gas upgrading. The cordierite monoliths of the
honeycomb catalyst are coated with γ-Al2O3 and nickel. Lab tests conducted with varying conditions
have shown the production of high-quality SNG, with a CH4 content of up to 68 vol.%. The membrane
technology improved gas quality by raising the CH4 content above 96 vol.% [73].
D7.7 Analysis on future technology options and on techno-economic optimization Page 43 of 89
According to Gruber et al., it is possible to reach 77.5 % efficiency with a Ni-catalyzed fixed bed
methanation reactor with an Al2O3-based support [51].
Besides improving the methanation process itself, it is also possible to improve the overall efficiency
of PtG systems by enhancing the implementation of the methanation process in the overall process
route. Kirchbacher et al. investigated one such possibility in 2018 by studying the direct integration
of raw biogas into a methanation- and membrane-based PtG application. In this study, the PtG ap-
plication was coupled with two-stage fermentation [75].
As shown in Figure 5-6, gas upgrading works well, and the requirements for the Austrian gas grid
have been reached [75].
Figure 5-6: Results of the gas upgrading system of Kirchbacher et al. for different biogas compositions [75]
5.3 CO2 separation
Absorption is currently the dominant CO2 separation method in industrial scale. However, membrane
technology is considered as a promising technology, but is still under development. The biggest
problem with membrane technology in the industrial scale is its insufficient long-term stability and
wetting under real operating conditions. Real gases consist of minor SOx, NOx, CO, and water con-
tent, which have a negative effect on the lifetime of polymeric membranes. Therefore, testing under
real conditions is essential to make such technologies feasible for industrial scale. The catalyst of
the methanation reactor has been based on Ni with Al2O3 as support (Meth 134® by C&CS). Thus,
the reactor temperature of the setup has always been above 250 °C. The membrane material has
been polyimide [76].
To commercialize the technology, the following factors have to be improved [76]:
Plasticization resistance
Thermal and chemical resistance
Long-term stability
Cost effectiveness
D7.7 Analysis on future technology options and on techno-economic optimization Page 44 of 89
Norahim et al. has given a short overview of current polymers in a review on membrane separation
technologies. It was found that polyimides perform better than the other polymeric membranes that
were tested. This can be seen in Table 5-1 [76].
Table 5-1: Overview of currently developed membrane separation technologies [76]
Material Testing conditions Performance
Polymer T [°C] p [bar] CO2:N2 pCO2 [Barrer] Selectivity
CO2/N2
Polysulfone RT 4 Single gas 0.71 1.61
Polysulfone 25 N/A Single gas 6 38
Matrimid® 9725 25 10 Single gas 6.2 27.5
Matrimid® 9725 35 9 50:50 4 23
Matrimid® 5218 30 2 10:90 8 27
Matrimid® 5218 (HF) 35 4 55:45 16 28
Matrimid® 5218/PES, 80/20 (HF) 35 4 55:45 30 36
6FDA-TMPDA 35 60 10:90 400 17
Pebax® MH-1657 25 5 Single gas 55.8 40.2
Pebax® MH-1657 25 2 10:90 480 48
Pebax® MH-1657 30 2 Single gas 60 57
Pebax® MH-1657 30 0.6 Single gas 73 45
Pebax® MH-1657/PEG 30 0.6 Single gas 151 47
Pebax-1074 25 3 Single gas 111 50
Polytherimide 25 1 Single gas 1943 2.3
Cellulose acetate N/A 3 Single gas 401 32.92
Table 5-1 also shows that, between Matrimid® and hexafluoro-substituted aromatic polyimides (6-
FDA), the latter perform better due to the greater free volume following from the bulky –C(CF3)2
group in it [76].
D7.7 Analysis on future technology options and on techno-economic optimization Page 45 of 89
6 Potential future fields of application
PtG is being promoted as a promising technology for providing chemical energy produced from re-
newable electric power in future low-carbon energy systems. The technology is known to provide
significant advantages in storage density and, in combination with methanation, even acts as a car-
bon sink by supporting the long-term fixation of CO2. Nevertheless, the process has conversion effi-
ciency issues; this energy efficiency question constitutes a major disadvantage compared to the
direct usage of renewable electric energy. Nevertheless, PtG is expected to be used in future renew-
able energy systems in certain applications. These future fields of PtG application are discussed
below, as we provide an overview of the most promising research activities and proposed solutions.
6.1 Energy storage and transportation
Even though state-of-the-art PtG applications display significant conversion losses due to efficiency
issues, they still have major advantages in achievable storage densities compared to other technol-
ogies, like the direct storage of electric energy in batteries (cf. Figure 6-1). PtG applications use
significantly less space for larger storage capacities, keeping the sealing of land areas to a minimum,
which is a desirable economic aspect. PtG also allows the long-term storage of excess energy in
widely applicable energy forms, while the discharge times of competing technologies are far more
limited, as shown in Figure 6-1. Additionally, the costs for storage and the transportation infrastruc-
ture required to handle increasing amounts of renewable power will be marginal, as the existing gas
infrastructure can be used as long as the produced gas meets the appropriate requirements.
Figure 6-1: Storage capacity of different energy storage applications
Source: Renewables global futures report [77], Fraunhofer Institute, Germany, 2014; edited by author
The injection of hydrogen, or rather hydrogen-enriched methane, into existing regional (and supra-
regional) gas grids and infrastructure has been intensively discussed in recent PtG considerations.
This issue is critical, as the national requirements for gas quality and composition differ between
gas-transiting countries. However, recent studies show that an increase of hydrogen content in the
existing infrastructure is not necessarily a problem for subsequent utilization [78].
D7.7 Analysis on future technology options and on techno-economic optimization Page 46 of 89
To allow centralized production and reduce efforts for transportation of energy TenneT plans to build
an artificial island in the middle of the Dogger Bank in the North Sea, which has an average depth
of 25 m and is therefore a promising location for offshore wind farms. This artificial island, dubbed
the “North Sea Wind Power Hub,” is planned to connect wind farms with power of up to 100 GW by
mid-century and provide the advantage of shorter power cables and reduced transport and mainte-
nance costs. It will later be connected to grids in the Netherlands, United Kingdom, Belgium, Ger-
many, Norway and Denmark [79,80]. A PtG plant could be used for storage of excess electricity
production in the North Sea Wind Power Hub.
The question of disused natural gas fields is important in discussions about reusing existing gas
infrastructure. Recent research [81,82] has shown that natural underground storage such as pore-
space storage can easily be used for hydrogen injection to enable long-term storage without signifi-
cant losses. As a positive side effect, investigations have shown that the presence of microbes in
these underground storage spaces causes the transformation of hydrogen gas to methane when
CO2 is injected. This would avoid the need for subsequent methanation steps as part of the PtG
plant. It is currently being investigated as part of Austria’s Underground Sun Conversion [83] re-
search project. If the results are promising in terms of efficiencies, conversion rates, efforts for injec-
tion, and the removal of product gases (gas purity), this approach may be able to significantly reduce
infrastructure costs for future PtG applications.
A similar approach was discussed by Jensen et al., who proposed the application of highly efficient
PtG plants in combination with the underground storage of CO2 and CH4. The research claims that
this would incur storage costs comparable to those for pumped hydro and much lower than those for
previously proposed technologies [84]. Besides the integration of existing natural storages, this ap-
proach benefits from the thermal coupling of high-temperature electrolysis and methanation to
achieve high round-trip efficiencies for electricity storage of 70% and beyond. The system is planned
to use electrical surpluses for water electrolysis based on reversible solid oxide cell technology. This
renewable hydrogen gas will then be chemically converted to SNG, while highly integrated thermal
management between endothermal electrolysis and exothermal methanation will be used to in-
crease overall electrical efficiency. The process can also be reverted by oxidizing the previously
stored SNG in a reversible solid oxide cell, producing electrical energy if required. CO2 from the
oxidation process is again stored in underground storage. A schematic of the process is shown in
Figure 6-2. [84]
This kind of application illustrates PtG’s potential as an energy-balancing method for handling excess
and peak loads in future energy systems, which are dominated by fluctuating renewable energy
sources like wind and solar power. By using high-temperature electrolysis in combination with inte-
grated exothermal processes (e.g., methanation) or by being supplied with renewable waste heat
from external sources, electrical conversion efficiencies above 85% are expected (cf. goals of the
HELMETH project [46]), which are almost competitive with the direct storage of electric power (e.g.,
in batteries), especially when discharge times are considered (cf. Figure 6-1). A similar objective is
being pursued by the Austrian HydroMetha research project [45]. In addition to the integrated thermal
coupling of electrolysis and methanation, the process includes a chemical reduction of CO2 to CO in
the electrolysis step. This so-called “co-electrolysis” leads to further performance gains: first, the
methanation of CO, instead of CO2, is higher exothermal, providing additional heat, which can be re-
coupled to the solid oxide electrolysis cell; second, the chemical conversion of CO2 within the elec-
trolysis cell at elevated temperatures (e.g., 800°C instead of 250°C) allows for higher efficiency due
to the reduced Gibbs Free Energy in the reaction [85]. Hence, overall conversion efficiency can be
increased.
D7.7 Analysis on future technology options and on techno-economic optimization Page 47 of 89
Figure 6-2: Schematic diagram of the rSOC electricity storage system [84]
6.2 Industrial processes
Due to its significant losses in energy conversion, PtG is often considered questionable as an energy
source for carbon-demanding processes that could be electrified in the intermediate future. Hence,
the electrification of individual processes for the direct usage of renewable electrical energy is the
best method, providing the highest primary energy efficiencies and decarbonization in future energy
systems dominated by renewables. However, PtG is still expected to play a valuable role in sectors
where a complete electrification of incorporated processes is not expected in the near future but
where the usage of fossil fuels and resources must be mitigated. In this context, PtG can serve as
an interim technology on the path from highly established to stepwise, adaptable industrial pro-
cesses, such as in steel production and chemical industries, where high amounts of fossil natural
gas are processed.
Besides this prolonged sustainable usage of sophisticated production and conversion processes,
highly efficient alternatives for almost non-electrifiable value chains have to be developed. In the
steel industry, the direct reduction of iron by using renewable hydrogen from water electrolysis has
been discussed in recent years. Various projects, such as H2Future [86], GrInHy [87], and HYBRIT
[88], are seeking to identify the process chains that could meet today’s standards.
The most common reduction agent in steelmaking is based on coal and therefore fossil resources.
Hence, this must be replaced by renewable alternatives. The projects mentioned above aim to use
product gases from a PtG application as a reduction agent. An example of a process scheme related
to the HYBRIT approach with its specific energy carrier flow is shown in Figure 6-3 [89].
D7.7 Analysis on future technology options and on techno-economic optimization Page 48 of 89
Figure 6-3: Scheme of the HYBRIT process flow and its energy balance [89]
In addition to the direct utilization of gases produced by PtG applications, particularly hydrogen and
SNG, there are many ways that these could be further processed to generate renewable hydrogen-
or carbon-based end-products. This process is commonly generalized under the term “Power-to-X”
(PtX). PtX can provide sustainable solutions for decarbonization and CCU. Besides the long-term
fixation of CO2 (e.g., in renewable polymers), renewable hydrogen will be important in chemical in-
dustry. The fertilizer industry, for example, is a major emitter of CO2 related to ammonia production,
mainly based on hydrogen production from natural gas [90]. Alternatives are currently being investi-
gated (cf., e.g. [91]).
6.3 Mobility
Another application of PtG, or rather PtX, is the generation of synthetic fuels, such as methane,
methanol, and ethanol [92], as substitutes for fossils in conventional internal combustion engines, or
using hydrogen directly in fuel cell electric vehicles (FCEV) [93]. Even though today’s focus is more
on battery electric vehicles (BEVs), the competiveness in terms of cruising range is only partly given.
Renewable gaseous and liquid fuels still outperform batteries in terms of energy density, as shown
in Figure 6-1. This is especially true for light and heavy-duty commercial vehicles like trucks and
buses, but also applies to shipping, which is highly dependent on fossil fuels.
Using hydrogen and carbon dioxide as base materials allows for the generation of a wide range of
potential fuels. Using SNG in the form of liquefied natural gas (LNG) could be useful, especially in
shipping [94]. An overview of potential synthetic fuel production routes is shown in Figure 6-4. Many
public transport buses are already powered by natural gas rather than diesel. Hence, switching to
SNG from renewable sources would not require additional adaptations or investments for vehicles.
D7.7 Analysis on future technology options and on techno-economic optimization Page 49 of 89
Figure 6-4: Overview of possible synthetic fuel production process routes [94]
For the direct usage of hydrogen in the mobility sector, energy conversion is performed in fuel cells
rather than in combustion engines. This avoids the conversion losses of methanation (or any other
fuel generation process), increasing overall energy efficiency.
Fuel cell-based drive systems are already available in many vehicles (cf. Figure 6-5). As mentioned,
utility vehicles (for transporting both humans and goods) are benefiting from the range and weight
advantages produced by the elevated energy densities (relative to batteries). Appropriate implemen-
tation for busses (cf. [95]) and trucks (cf. [96]) have been investigated and tested in various inde-
pendent studies and pilot projects for fleets of up to medium size. Hence, hydrogen is a promising
renewable mobility and fuel option for long-range fleets (e.g., in the transport sector), with short dwell
times. [97]
Figure 6-5: Overview of BEV and FCEV application fields [97]
(HEV…hybrid electric vehicle, PHEV…plug-in hybrid electric vehicle)
D7.7 Analysis on future technology options and on techno-economic optimization Page 50 of 89
7 Economic evaluation
The many potential applications of PtG in the energy system may lead to various functions and
benefits, from which numerous fields of application can be derived. The economic evaluation and
analysis of the optimal plant configuration below is based on the specific production costs for SNG.
The following fields of application are investigated:
1. PtG plant powered by a photovoltaic power plant (PtG-PV)
2. PtG plant powered by a wind farm (PtG-Wind)
3. PtG plant powered by the public grid (PtG-Grid)
The calculations and analyzes of SNG production costs are carried out using PResTiGE (Power-to-
Gas Assessment Tool) developed at the Energieinstitut an der JKU Linz (see Figure 7-1). PResTiGE
is a toolbox for current and prospective techno-economic and environmental benchmarking of PtG
systems. The EXCEL tool comprises data from demo sites and benchmark systems as options for
electricity storage or applications of the gaseous products H2 or CH4 in the transportation sector at
different scales, in forms that are regionally adaptable over all process steps of the PtG system and
product application. The assessment results reveal the optimal PtG system configuration and imple-
mentation (i.e., with minimal cost and maximal system benefits). Sensitivities can be systematically
analyzed to explore the robustness of the results.
Figure 7-1: Overview of the tool PResTiGE
The quantitative economic assessment via PResTiGE is based on the specific production costs of
hydrogen or SNG, which are calculated from the total annual costs in relation to the amount of an-
nually produced energy. The total annual costs are calculated using the so-called “annuity method”
following VDI 2067.
In calculating SNG production costs, an interest rate of 4% and a period of 20 years are assumed.
No price change factor is taken into account. The total annual costs include capital- , demand- ,
operating-related, and other costs. The capital-related costs are investment and replacement costs.
Annual demand-related costs include energy costs and costs for auxiliary energy. Operating-related
D7.7 Analysis on future technology options and on techno-economic optimization Page 51 of 89
costs include annual costs for the maintenance, operation, and cleaning of the plant. Other cost
items include insurance, levies, and administration costs. The calculated SNG production costs do
not include taxes and charges or any electricity and gas network tariffs, as these depend on the
country in which the PtG plant is being built.
The following chapters describe the most important input parameters for calculating SNG production
costs, such as electricity costs and quantities, investment costs, efficiency of the PtG plant, lifetime
of the electrolyzer, costs for CO2, revenue from the utilization of waste heat and oxygen, and hot
standby power consumption. The SNG production costs for different technologies and time horizons
are also calculated, and a sensitivity analysis is carried out.
7.1 Electricity costs and quantities
The electricity costs and power available for the plant are crucial for both the economic assessment
and plant operation. These input parameters vary depending on the field of application. Likewise,
the electricity market is subject to constant change. This chapter describes the electricity procure-
ment scenarios, including prices and quantities.
7.1.1 Electricity from photovoltaic power plant or wind farm
In the PtG-PV and PtG-Wind scenarios, the PtG plant obtains electricity directly from a photovoltaic
power plant or wind farm. Two options are investigated to analyze the different operation modes:
1. Use of the total electricity generated. Here, the PtG plant is designed for the maximum output
of PV/wind (PtG-PV-100% and PtG-Wind-100% scenarios)
2. Use of a part of the electricity generated. Here, a certain amount of the power produced from
PV/wind is fed into the public grid. The remaining produced electricity is used to produce
SNG by the PtG plant (PtG-PV-50%, PtG-PV-75%, PtG-Wind-50%, and PtG-Wind-75% sce-
narios)
When operating the PtG plant with electricity from a wind farm or photovoltaic power plant, only the
directly generated power can be used in the electrolyzer. Figure 7-2 and Figure 7-3 show the typical
electricity production characteristics of a wind farm and photovoltaic power plant. The generation
profiles are based on the electricity production of a wind farm (14 MW) and PV power plant (about
3 MWp) in Austria, which were scaled up to a maximum power of 100 MW for comparison. The ana-
lyzed operating modes differ in the share of power fed into the grid, which ranges from 0% (no feed-
in) to 50% (half of the power is fed into the grid) of the nominal power of the production plant. The
remaining share of the produced power is used by the PtG plant. For example, in Figure 7-2 and
Figure 7-3, the maximum grid feed-in is 25% of the maximum power of the wind farm/PV power
plant. The remaining power, 75% of the maximum power, is used by the PtG plant to produce SNG.
In a future energy system with a high proportion of PV and wind power, this operation mode could
be used to feed urgently needed electricity into the grid as a kind of base load. In addition, renewable
gas could be produced with the electricity power peaks that would otherwise severely burden the
grid and lead to grid expansion. The renewable gas serves as a long-term energy storage (transfer
energy from the summer to the winter) or as an energy source in industry and mobility applications.
D7.7 Analysis on future technology options and on techno-economic optimization Page 52 of 89
Figure 7-2: Typical electricity production characteristic of a 100 MW wind farm. The power is divided in the share of grid
feed-in (green coloured) and power for the PtG-plant (orange coloured). In this example is the maximum grid feed-in 25 % of the maximum power of the wind farm (25 MW), the electrolyzer (75 MW) uses the surpluses
Figure 7-3: Typical electricity production characteristic of a 100 MW photovoltaic power plant (app. 123 MWp). The
power is divided in the share of grid feed-in (green colored) and power for the PtG-plant (blue colored). In this example is the maximum grid feed-in 25 % of the maximum power of the photovoltaic plant (25 MW), the electrolyzer (75 MW) uses
the surpluses
The different generation characteristics of the wind farm and the PV power plant are shown in the
ordered annual production curve, where the ordered power values are presented over the year,
indicating the frequency of the power over the course of a year (see Figure 7-4). The area under the
production curve reflects the energy produced.
The wind farm runs at a rated power for a longer time than the PV power plant does. The PV power
plant generates at only about 4,300 h/a energy (no production at night), while the wind farm gener-
ates over 7,700 h/a. Thus, a 100 MW wind farm generates significantly more energy than a 100 MW
PV system.
D7.7 Analysis on future technology options and on techno-economic optimization Page 53 of 89
Figure 7-4: Ordered annual production curve of a 100 MW wind farm and PV power plant
The generation characteristics of the renewable energy plant strongly influence the operation and
economic efficiency of the PtG plant. Key data on the wind farm and PV power plant are summarized
in Table 7-1.
Table 7-1: Key data of the wind farm and PV power plant
Characteristic Unit Wind farm PV power plant
Maximum power MW 100 100
Annual energy production MWh 310,056 139,483
Full-load hours h/a 3,101 1,136
Time power = 0 h/a 1,018 4,473
Time power > 0 h/a 7,742 4,287
Electricity procurement costs PV
According to [98], the levelized costs for electricity (LOEC) from PV power plants were about 54 to
84 €/MWh in 2014 depending on the location. In 2050, the LOEC are estimated to be in a range of
25 to 44 €/MWh in southern Germany (at 1,190 kWh/kWp) and 18 to 31 €/MWh in southern Spain
(at 1,680 kWh/kWp). A more recent study [99] claims that the current costs (2018) for electricity from
large PV power plants in Germany are in a range of 37.1 to 67.7 €/MWh. The estimated costs for
2035 are in a range of 21.6 to 39.4 €/MWh. A more conservative estimation of electricity costs from
PV in 2050, of about 55 to 84 €/MWh, is provided in the EU Reference Scenario [100]. However, this
estimation seems already outdated, since the current costs are lower.
Following the literature, this Deliverable assumes electricity costs from large-scale PV power plants
at locations with an average solar radiation of 40 €/MWh, 30 €/MWh, and 20 €/MWh for 2020, 2030,
and 2050, respectively.
Electricity procurement costs for wind
In [101], the generation costs for onshore wind farms in Germany are estimated at about 30–
60 €/MWh and 25–50 €/MWh in 2030 and 2050, respectively. In a more optimistic scenario, where
a higher potential for cost reduction is taken into account, the electricity costs from wind power de-
cline to 25–45 €/MWh and 20–35 €/MWh, respectively. The average electricity costs for projects in
Germany in 2016/17 for an average-quality site was about 65 €/MWh. When compared to other parts
of the world (e.g., Morocco, Peru, Mexico), the costs for electricity from onshore wind farms are even
D7.7 Analysis on future technology options and on techno-economic optimization Page 54 of 89
lower, as shown in wind auctions, where the average bids were in a range of 27–34 €/MWh. The
average bids at wind auctions for offshore wind farms are higher, at about 50–72 €/MWh. According
to [99], the 2018 costs for electricity from onshore wind farms in Germany were in a range of 40–
82 €/MWh, depending on the location. Although the offshore wind farms have higher full-load hours,
the costs for electricity are higher, at about 75–138 €/MWh. In the long term, until 2035, the costs for
electricity from onshore wind turbines in Germany will decline to 35–71 €/MWh, and the costs for
offshore wind turbines will decline to 57–101 €/MWh. Analogous to the cost estimation for electricity
from PV power plants, the costs for electricity from wind turbines in 2050 are overestimated in the
EU Reference Scenario at 72–90 €/MWh [100], as the costs are lower today.
Following the literature, this Deliverable assumes average costs for electricity form wind power plants
(onshore as well as offshore) of about 60 €/MWh, 50 €/MWh, and 40 €/MWh for 2020, 2030, and
2050, respectively.
7.1.2 Electricity from the spot market
Based on 2017 spot market prices for electricity with a time resolution of 15 minutes in Austria,
forecasts for 2020, 2030, and 2050 are prepared. The forecasts depend on the development of the
average spot market price and price volatility.
For the grid-connected 100 MWel PtG plant, it is assumed that an ideal power grid provides the
power. This means that the required power is available at any given time. The investigated operating
modes differ in terms of the full-load hours (from 1,000–8,000 h/a) of the plant.
Average spot market prices
The average spot market price in 2017 was about 34.5 €/MWh and fluctuated between a minimum
price of -102 €/MWh and a maximum of 170 €/MWh (see Figure 7-5) [102].
The EU Reference Scenario 2016 forecast average electricity prices before taxes for households,
services, and industry until 2050. The price consists of annual capital costs, fixed costs, variable
costs, fuel costs, taxes on fuels and ETS payments, and grid costs. For the industry sector, the price
remains quite stable, ranging from 90 to 100 €2013/MWh from 2010 to 2050. It is expected to reach
about 97 €2013/MWh by 2030 and about 99 €2013/MWh by 2050 [100].
The 2014 study in [103] forecasts the development of wholesale electricity prices until 2050. The
analysis assumes that prices will decline until 2020, mainly due to the priority feed-in of renewable
energy. The subsequent increase in prices is explained as being caused by rising costs for CO2
emissions and fuel prices, as well as the effects of the nuclear phase-out. The demand for new
capacity to secure peak load coverage also increases. The estimated wholesale electricity prices in
Germany are 67 €2011/MWh and 87 €2011/MWh for 2030 and 2050, respectively.
The Sustainable River Management study [104] examines the possible long-term development of
spot market electricity prices in Austria on the basis of two independent electricity price models (ewi
Energy Research & Scenarios GmbH and enervis energy advisor GmbH) for 2025, 2035, and 2050.
According to the enervis model, the electricity prices are 56.0 €2016/MWh, 75.7 €2016/MWh, and
75.4 €2016/MWh for 2025, 2035 and 2050, respectively. The prices calculated with the ewi model are
similar, at 56.9 €2016/MWh, 75.8 €2016/MWh, and 78.6 €2016/MWh for 2025, 2035, and 2050, respec-
tively.
The EU Energy Outlook 2050 report examines long-term trends in the European energy system,
including developments in electricity prices. The development of average, unweighted electricity
prices until 2040 heavily depends on the primary energy and CO2 prices. From 2040, electricity
D7.7 Analysis on future technology options and on techno-economic optimization Page 55 of 89
prices will stagnate despite increasing gas and CO2 prices due to high wind and photovoltaic feed-
ins, which are increasingly leading to low (and often negative) electricity prices. Future developments
depend on the expansion of renewable energies and will therefore vary across countries. In 2030,
the electricity price is expected to be about 70 €/MWh; in 2050, the price will rise to about 80 €/MWh
(in a range from 70–120 €/MWh depending on the amount of RES installed) [105].
In a Greenpeace scenario regarding an energy concept for Germany, electricity prices (base spot
market price) are estimated to be much lower, at about 45 €/MWh in 2030 and 22 €/MWh in 2050,
than in other studies on the development of electricity prices. This is justified by the fact, that starting
from 2030, power plants with small marginal costs are price-determining. These include increasingly
renewable energy plants with marginal costs close to zero and the remaining coal-fired power plants.
As renewable energy use expands, the use of conventional power plants will become increasingly
rare, which will bring the electricity costs closer to those of renewable energies [106].
[107] analyzes how implementing 65% of renewable energy until 2030 and a gradual phase-out of
coal power generation in Germany will affect electricity prices, CO2 emissions, and the electricity
market. In three different scenarios, stock market electricity prices of 53–61 €/MWh are estimated
for 2030. The development of the electricity prices is mainly dependent on the assumption of in-
creasing CO2 emissions and fuel prices. The accelerated phase-out of coal power generation would
lead to an increase from 57 €/MWh to 61 €/MWh. With a simultaneously accelerated expansion of
renewables to 65% in 2030, the costs would decrease to 53 €/MWh.
The study of [108] on the climate-protection contribution of the power sector until 2040 also forecast
wholesale electricity prices. In this study, electricity prices also rise due to the assumed increase in
coal and gas costs. It is estimated that prices will rise until about 2028 and then stabilize in a range
of 60–65 €/MWh.
Table 7-2 summarizes the electricity prices forecasted for 2030 and 2030 in the analyzed sources.
Table 7-2: Forecast for electricity prices in 2030 and 2050
2030
€/MWh
2050
€/MWh Source
97 99 [100]
67 87 [103]
56.0/75.7*
56.9/75.8*
75.4
78.6 [104]
70 80 (70–120) [105]
45 22 [106]
53–61 - [107]
60-65 - [108]
66 74 Mean value
66 79 Median value
* … values for the year 2025/2035
D7.7 Analysis on future technology options and on techno-economic optimization Page 56 of 89
For the further calculations in this report, the median value (where outliers have less priority) in the
examined studies was assumed to be approximately 65 €/MWh for 2030 and 80 €/MWh for 2050.
For 2020, the average spot market prices observed in 2017 were assumed (around 35 €/MWh).
For the estimation of future spot market prices, the price volatility for 2030 and 2050 will also be
changed, in addition to the average prices. Both volumes and prices will likely be more volatile due
to the expanding renewable power generation. Based on the literature review, a mean spot market
price of 65 €/MWh for 2030 and 80 €/MWh for 2050 is assumed (see Table 7-2). The rising volatility
is taken into account by multiplying the deviation from the mean value occurring in one hour by a
volatility factor. This factor is set to 1.5 (+50%) for 2030 and 2 (+100%) in 2050. The resulting trends
in spot market prices over the year are shown in Figure 7-5. The figure shows higher spot market
prices in 2030 and 2050 and higher volatility in the form of significantly larger fluctuations in spot
market prices compared to the 2017 reference data.
Figure 7-5: Spot market prices for 2017 and forecasts for 2020, 2030 and 2050
Fluctuations in spot market prices are taken into account in the economic evaluation of the operating
modes of the PtG plant. Depending on the specific production costs, the cost-optimal full-load hours
are determined for the respective application. The amounts of electricity, and thus the capacity utili-
zation of the PtG plant, are dependent on the electricity procurement costs.
It is assumed that the PtG system is always operated at times with the cheapest electricity prices.
This means that, when operating the system with certain full-load hours, electricity may be purchased
only up to a certain price. This situation results in a separate average electricity price for each full-
load hour (see Table 7-3).
If, for example, a PtG plant in 2050 is operated only when the spot market prices are below
69 €/MWh, the plant will reach 3,000 full-load hours, and the average electricity purchase price will
be around 16.2 €/MWh. This value is below the average spot market price of 80 €/MWh in 2050,
since only the cheapest hours of the year are used. When operating the PtG plant with higher full-
load hours (e.g., 8,000 h/a), the mean electricity purchase price (66.7 €/MWh) converges with the
average spot market price in 2050 (80 €/MWh), as almost all hours, even those with higher prices,
have to be used to achieve the required operating time.
D7.7 Analysis on future technology options and on techno-economic optimization Page 57 of 89
Table 7-3: Mean prices for electricity from the spot market for certain full load hours in 2020, 2030 and 2050